WO2012026382A1

WO2012026382A1 - Magnetic resonance imaging device and vibrational error magnetic field-reducing method

Info

Publication number: WO2012026382A1
Application number: PCT/JP2011/068650
Authority: WO
Inventors: 昇一宮脇; 竹内　博幸
Original assignee: 株式会社日立メディコ
Priority date: 2010-08-26
Filing date: 2011-08-18
Publication date: 2012-03-01
Also published as: JPWO2012026382A1; JP6038654B2; US20130147481A1; US9453897B2

Abstract

In order to reduce, irrespective of measurement conditions, the deterioration of image quality arising from vibrational error magnetic fields generated by vibration of the mechanical structures of MRI devices, error magnetic field image data representing the error magnetic field distribution is acquired on the basis of echo signals measured using pulse sequences having a test gradient magnetic field. Using the error magnetic field image data, parameter values for a damped vibration function representing the vibrational error magnetic field are calculated and a correcting magnetic field is calculated on the basis of the parameter values for the damped vibration function representing the calculated vibrational error magnetic field.

Description

Magnetic resonance imaging apparatus and vibration error magnetic field reduction method

The present invention relates to a magnetic resonance imaging (hereinafter referred to as MRI) apparatus that obtains a tomographic image of an arbitrary part of a subject by utilizing a nuclear magnetic resonance (hereinafter referred to as NMR) phenomenon, and in particular, an MRI apparatus structure generated by applying a gradient magnetic field The present invention relates to a technique for correcting a vibration error magnetic field caused by vibration of an object.

The MRI device measures NMR signals generated by the spins of the subject, especially the tissues of the human body, and visualizes the form and function of the head, abdomen, limbs, etc. in two or three dimensions Device. In imaging, the NMR signal is given different phase encoding depending on the gradient magnetic field and is frequency-encoded and measured as time-series data. The measured NMR signal is reconstructed into an image by two-dimensional or three-dimensional Fourier transform.

In the MRI apparatus, when imaging is performed based on a predetermined pulse sequence, it is necessary to selectively excite a specific region and to accurately and arbitrarily control the application time and intensity of the gradient magnetic field. However, when a gradient magnetic field is generated, a damping current is induced in the conductive structure around the gradient coil. This is called eddy current (Eddy Current) and generates a magnetic field that changes in space and time. As a result, the gradient magnetic field perceived by the subject deviates from the ideal state and appears as various image quality degradations such as image distortion, signal strength reduction, and ghost. Various methods for correcting the error magnetic field caused by the eddy current have been studied and many patents have been proposed.

On the other hand, when a gradient magnetic field is generated in the MRI apparatus, Lorentz force due to the gradient magnetic field coil current spreads to surrounding structures. Generation and extinction of a gradient magnetic field in MRI imaging has a specific direction, and is frequently and periodically performed. Therefore, a force that spreads to surrounding structures also has directionality and periodicity. In other words, while performing MRI imaging, it can be said that the MRI apparatus is always “vibrated” in some direction and cycle. A particular problem arises when the direction and period of excitation match the natural vibration characteristics determined by the mechanical structure of the MRI apparatus. It has been found that an oscillating error magnetic field generated by a resonance phenomenon of a mechanical structure has a non-negligible influence on image quality. Conventional MRI devices have been designed to prevent this resonance phenomenon from occurring or to have a mechanical structure that does not affect the image quality even if it occurs. There is a need for a design that allows the generation of resonance.

One example of a technique for avoiding a vibration error magnetic field due to vibration of this mechanical structure is Patent Document 1. In the technique presented in Patent Document 1, natural vibration information of a target MRI apparatus is measured and analyzed in advance using a reference gradient magnetic field waveform and stored. In actual MRI imaging, the vibration of the mechanical structure generated by the execution of the pulse sequence is estimated based on the measurement conditions set by the operator and the natural vibration information prepared in advance. Then, it is determined whether or not the estimated value of the vibration exceeds the allowable amount, and if it exceeds, the change of the measurement condition is prompted.

JP 2005-270326 A Japanese Patent No. 4106053 U.S. Pat. International Publication WO2004 / 004563 International Publication WO2010 / 143586

However, considering that the combinations of gradient magnetic field waveforms determined by the measurement conditions are infinite, it is difficult to accurately estimate all the vibrations of mechanical structures that actually occur based on the natural vibration information prepared in advance. It is done. In addition, there is a waiting time required for internal processing (calculation) to predict the vibration to occur before the operator starts measurement, and it is necessary to change the imaging conditions if it is inappropriate measurement conditions. Not preferable. Furthermore, given that the FSE 振動 (Fast Spin Echo) sequence, which accounts for the majority of clinical measurements, is significantly affected by the vibration error magnetic field, many problems remain to be solved before this technology can be put to practical use. It is thought that there is.

Therefore, the present invention has been made in view of such a problem, and it is possible to reduce image quality degradation caused by a vibration error magnetic field generated by vibration of a mechanical structure of an MRI apparatus regardless of measurement conditions. An object of the present invention is to provide a possible MRI apparatus and vibration error magnetic field reduction method.

In order to achieve the above object, the present invention acquires error magnetic field image data representing an error magnetic field distribution based on an echo signal measured using a pulse sequence having a test gradient magnetic field. The parameter value of the damped vibration function representing the vibration error magnetic field is calculated using the data, and the correction magnetic field is calculated based on the parameter value of the damped vibration function representing the calculated vibration error magnetic field. This parameter value is a characteristic value representing the vibration characteristic of each MRI apparatus.

Specifically, the MRI apparatus of the present invention is caused by the application of a gradient magnetic field, a static magnetic field generation unit that generates a static magnetic field in an imaging space, a gradient magnetic field generation unit that generates a gradient magnetic field superimposed on the static magnetic field, and a gradient magnetic field. A correction magnetic field generation unit that generates a correction magnetic field that corrects an error magnetic field generated in the imaging space, a static magnetic field generation unit, a gradient magnetic field generation unit, and a correction magnetic field generation unit that are installed to support these, and a predetermined structure Measurement control unit that measures echo signals from the subject placed in the imaging space based on the pulse sequence of the above, and correction magnetic field calculation to obtain a correction magnetic field for correcting the error magnetic field generated in the imaging space due to the application of the gradient magnetic field The correction magnetic field calculation unit calculates an error magnetic field including an oscillation error magnetic field based on the vibration of the structural unit caused by application of the gradient magnetic field, and corrects the calculated error magnetic field. Seeking The features.

Further, the vibration error magnetic field reduction method of the present invention includes a measurement step for measuring an echo signal using a pulse sequence having a test gradient magnetic field, and an error magnetic field image representing an error magnetic field distribution for each sampling time using the echo signal. A step of acquiring data, a parameter value calculating step of calculating a parameter value of a damped vibration function representing a vibration error magnetic field using error magnetic field image data for each sampling time, and a damping vibration function representing the calculated vibration error magnetic field A correction magnetic field calculation step for calculating a correction magnetic field based on the parameter value.

According to the MRI apparatus and the vibration error magnetic field reduction method of the present invention, it is possible to reduce image quality degradation caused by the vibration error magnetic field generated by the vibration of the mechanical structure of the MRI apparatus regardless of the measurement conditions.

Furthermore, this effect can improve the degree of freedom in the mechanical structure design of the MRI apparatus, and can allow a certain amount of vibration, so that the material cost of the MRI apparatus can be reduced. There is also a secondary effect.

The block diagram which shows the whole structure of an example of the MRI apparatus in this invention Functional block diagram of the corrected magnetic field calculation unit 200 6 is a flowchart showing a processing flow in which functional units of the correction magnetic field calculation unit 200 cooperate to correct the vibration error magnetic field. The figure which shows a suitable pulse sequence for the measurement of the high frequency component of a vibration error magnetic field. FIG. 5 is a diagram illustrating an example of an order in which Scan (+) and Scan (−) are performed for each phase encode value for measuring an echo signal in the pulse sequence of FIG. 4. The flowchart which shows the processing flow which acquires time series phase image data. The flowchart which shows the processing flow of a vibration error magnetic field analysis. (a) Tukey Window, (b) Example of vibration error magnetic field (X to X) multiplied by Tukey Window. FIG. 4A shows an example of an absolute spectrum, and FIG. 4B shows an example of a phase spectrum. An example graph of phase calibration data. The image which shows the effect of this invention. (a) is an image obtained when the vibration error magnetic field reduction processing of the present invention is not performed, and (b) is an image obtained when the vibration error magnetic field reduction processing of the present invention is performed.

Hereinafter, preferred embodiments of the MRI apparatus of the present invention will be described in detail with reference to the accompanying drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiments of the invention, and the repetitive description thereof is omitted.

First, an overall outline of an example of an MRI apparatus according to the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing the overall configuration of an embodiment of an MRI apparatus according to the present invention. This MRI apparatus uses a NMR phenomenon to obtain a tomographic image of a subject, and as shown in FIG. 1, a static magnetic field generation system 2, a gradient magnetic field generation system 1, a transmission system 3, and a reception system 5 A signal processing system 7, a measurement control unit 6, a calculation processing unit 8, and a correction magnetic field calculation unit 200.

The static magnetic field generation system 2 generates a uniform static magnetic field around the subject 9 in the direction of the body axis or in a direction perpendicular to the body axis. A magnetic field generating means (not shown) of a magnet system, a normal conducting system, or a superconducting system is arranged.

The gradient magnetic field generating system 1 is composed of a gradient magnetic field coil 10 wound in three axial directions of X, Y, and Z, and a gradient magnetic field power source 11 for driving each coil. By driving the gradient magnetic field power supply 11 of each coil, gradient magnetic fields Gs, Gp, and Gf in the three axial directions of X, Y, and Z are applied to the subject 9.

When imaging a two-dimensional slice plane, a slice gradient magnetic field pulse (Gs) is applied in a direction orthogonal to the slice plane (imaging cross section) to set a slice plane for the subject 9, and is orthogonal to the slice plane and orthogonal to each other Phase encoding gradient magnetic field pulse (Gp) and frequency encoding (leadout) gradient magnetic field pulse (Gf) are applied in the remaining two directions, and position information in each direction is encoded in the NMR signal (echo signal). .

In addition, a shim coil that forms part of the static magnetic field generation system 2 from the spatial and temporal information of the error magnetic field due to the eddy current and residual magnetic field due to the application of the gradient magnetic field, or the residual magnetic field, Each error magnetic field is reduced by applying a correction current to the local coil or the gradient magnetic field generation system 1.

The transmission system 3 irradiates a high-frequency magnetic field (hereinafter referred to as RF) pulse in order to cause an NMR phenomenon to atomic nuclei constituting the biological tissue of the subject 9, a high-frequency oscillator 12, a modulator 13, It comprises a high frequency amplifier 14 and an RF transmission coil 15 on the transmission side. Specifically, the high-frequency oscillator 12 is driven in accordance with a command from the measurement control unit 6 described later to generate a high-frequency pulse, the high-frequency pulse is amplitude-modulated by the modulator 13, and amplified by the high-frequency amplifier 14, and then the subject 9 The RF pulse is supplied to the RF transmission coil 15 arranged in the vicinity of the object 9 so that the subject 9 is irradiated with the RF pulse.

The receiving system 5 detects an echo signal (NMR signal) emitted by the NMR phenomenon of the nuclei constituting the biological tissue of the subject 9, and includes a receiving-side RF receiving coil 16, an amplifier 17, and quadrature detection. And an A / D converter 19. The response electromagnetic wave (NMR signal) of the subject 9 due to the electromagnetic wave irradiated from the RF transmission coil 15 is detected by the RF reception coil 16 arranged close to the subject 9 and passes through the amplifier 17 and the quadrature detector 18. Is input to the A / D converter 19 and converted into a digital quantity, and further, two series of collected data sampled by the quadrature phase detector 18 at the timing according to the command from the measurement control unit 6, and the signal is subjected to signal processing. It is to be sent to system 7.

The signal processing system 7 performs image reconstruction calculation and image display using the echo signal detected by the receiving system 5, and processes and measurement control such as Fourier transform, correction coefficient calculation, image reconstruction, etc. for the echo signal A processing unit 8 for controlling the unit 6, a ROM (read-only memory) 20 for storing a program for performing image analysis processing and measurement over time, an invariant parameter used in the execution, and the measurement obtained in the previous measurement A RAM (temporary writing / reading memory) 21 for temporarily storing parameters and echo signals detected by the receiving system 5 and an image used for setting the region of interest and storing parameters for setting the region of interest, and an arithmetic processing unit 8 Are read from the magneto-optical disk 22 and the magnetic disk 24 as data storage units for recording the image data reconstructed in It comprises a display 23 which is a display unit for converting the projected image data into an image and displaying it as a tomographic image.

The measurement control unit 6 is a control unit that repeatedly applies an RF pulse and a gradient magnetic field pulse based on a predetermined pulse sequence, and controls the measurement of an echo signal from the subject 9. Various commands necessary for data acquisition of tomographic images of the subject 9 are transmitted in the control system 3, shim coils and localized coils forming part of the static magnetic field generation system 2, the gradient magnetic field generation system 1, And to the receiving system 5.

The operation unit 4 is used to input control information for processing performed by the signal processing system 7, and includes a trackball, a mouse 25, and a keyboard 26. The operation unit 4 is disposed in the vicinity of the display 23, and an operator interactively controls various processes of the MRI apparatus via the operation unit 4 while looking at the display 23.

Currently, the radionuclide to be imaged by the MRI apparatus is a hydrogen nucleus (proton) which is the main constituent material of the subject, as is widely used in clinical practice. By imaging information on the spatial distribution of proton density and the spatial distribution of relaxation time in the excited state, the form or function of the human head, abdomen, limbs, etc. is imaged two-dimensionally or three-dimensionally.

As the background to the creation of the present invention, the inventor found that the measurement result of the error magnetic field caused by eddy current (hereinafter referred to as eddy current error magnetic field) includes an error caused by vibration of the mechanical structure of the MRI apparatus. This is due to the fact that the magnetic field component (hereinafter referred to as vibration error magnetic field) is also superimposed. The conventional (known) eddy current error magnetic field measurement method measures the response of the MRI device to the pre-applied test gradient magnetic field, so the measurement result includes not only the eddy current error magnetic field but also the vibration caused by the test gradient magnetic field. It should include the error magnetic field.

Therefore, the inventor recalled the MRI apparatus and vibration error magnetic field reduction method of the present invention that correct both the eddy current error magnetic field generated by applying the gradient magnetic field and the vibration error magnetic field caused by the vibration of the MRI apparatus structure. Next, one embodiment will be described in detail.

First, each function of the correction magnetic field calculation unit 200 that calculates and outputs a correction magnetic field for correcting the vibration error magnetic field according to the present embodiment will be described based on the functional block diagram of the correction magnetic field calculation unit 200 illustrated in FIG. . The correction magnetic field calculation unit 200 according to the present embodiment includes an error magnetic field measurement unit 201, an error magnetic field image acquisition unit 202, an error magnetic field calculation unit 203, an error magnetic field correction unit 204, a correction magnetic field calculation unit 205, and a phase calibration. Part 206. Each of these units is mounted in the measurement control unit 6 or the arithmetic processing unit 8. Then, these units cooperate to perform processing for correcting the vibration error magnetic field shown in the flowchart of FIG. Hereinafter, the outline of the vibration error magnetic field correction processing of the present embodiment will be described based on the flowchart of FIG.

In step 301, the vibration error magnetic field is measured. The error magnetic field measurement unit 201 generates a predetermined measurement sequence, causes the measurement control unit 6 to execute the measurement of the echo signal on which the vibration error magnetic field is superimposed, and outputs the echo signal on which the vibration error magnetic field is superimposed. get. Details will be described later.

In step 302, time-series error magnetic field image data is acquired using the echo signal acquired in step 301. The error magnetic field image acquisition unit 202 Fourier transforms the echo signal acquired in step 301 in the spatial axis direction, acquires a complex image for each sampling time of the echo signal, obtains a phase image from each complex image, Obtain time-series phase image data. Further, time-series error magnetic field image data is acquired from the time-series phase image data. Details will be described later.

In step 303, an oscillating magnetic field analysis is performed using the time-series phase image data acquired in step 302. Details of the oscillating magnetic field analysis will be described later.

In step 304, using the result of the vibration error magnetic field analysis in step 303, a correction magnetic field for correcting the vibration error magnetic field according to the input gradient magnetic field waveform is calculated. Details of the calculation of the correction magnetic field will be described later.

In step 305, phase correction of the correction magnetic field calculated in step 304 is performed. Details of the phase calibration will be described later.
The above is the description of the processing flow of the present embodiment. Details of each step will be described below.

(1. Vibration error magnetic field measurement)
Next, details of the vibration error magnetic field measurement in step 301 will be described.

The vibration error magnetic field caused by the vibration of the MRI apparatus structure induced by applying the gradient magnetic field has a frequency distribution. Therefore, in order to correctly measure the target frequency component, it is necessary to measure an echo signal on which a vibration error magnetic field is superimposed using a pulse sequence having an optimal time resolution and a measurement window (sampling time). Therefore, a known measurement method that satisfies this requirement, a method using a measurement sequence suitable for measuring a low frequency component of a vibration error magnetic field and a method using a measurement sequence suitable for measuring a high frequency component will be described below.

In any of the following methods, in the measurement of the vibration error magnetic field, an echo signal on which only the information on the vibration error magnetic field is superimposed in a state where the eddy current error magnetic field is corrected in advance is measured to obtain information on only the vibration error magnetic field. Alternatively, the eddy current error magnetic field may not be corrected, and an echo signal in which information on the eddy current error magnetic field and the vibration error magnetic field are superimposed may be measured, and the information may be acquired simultaneously. If you get both pieces of information at the same time,
The eddy current error magnetic field and the vibration error magnetic field are corrected without distinction.

(1.1 Measurement of low frequency components)
In order to measure a component having a low frequency of about 10 to 20 [Hz] or less, it is necessary to measure the error magnetic field fluctuation over a sufficiently long time. For this, the measurement method disclosed in Patent Document 2 is suitable. This method has no limitation on the measurement window (measurement time), so even if the error magnetic field fluctuation frequency is almost zero or the decay time is very long, the error magnetic field fluctuation can be measured significantly. It is.

In order to exclude the influence of error magnetic field due to gradient magnetic field for measuring echo signal and the influence of static magnetic field inhomogeneity, between two measurements with reversed polarity of test gradient magnetic field or with test gradient magnetic field The difference between the echo signal or the image after the Fourier transform may be acquired between the two measurements without and.

However, the method of Patent Document 2 has an upper limit on the frequency that can be measured because the time resolution is determined by the repetition time TR. Therefore, it is desirable to use the method described later for high frequency components.

The error magnetic field measurement unit 201 obtains data that specifically defines the application timing and application intensity of the gradient magnetic field pulse including the RF pulse and the test gradient magnetic field that constitute the pulse sequence described in Patent Document 2, or the sampling timing. To generate the pulse sequence. Then, the obtained data is notified to the measurement control unit 6, and the measurement control unit 6 is caused to execute a pulse sequence to measure an echo signal in which the vibration error magnetic field is reflected. In addition, when performing two measurements with the polarity of the test gradient magnetic field reversed or two measurements with and without the test gradient magnetic field, the error magnetic field measurement unit 201 generates each pulse sequence to generate a measurement control unit. 6 is executed respectively to measure echo signals having different test gradient magnetic fields, and acquire the difference between the echo signals or the images after Fourier transform thereof.

(1.2 Measurement of high frequency components)
One method for measuring high-frequency components is a technique for measuring echo signals by applying a test gradient magnetic field in advance and performing high-frequency excitation immediately after or after a predetermined time, as shown in Patent Document 3. It is done.

As another method, a measurement sequence as shown in FIG. 4 may be used. This measurement sequence is based on the Spin Echo sequence. RF, Gs, Gp, Gf, Echo, and A / D mean axes of RF pulse, slice gradient magnetic field, phase encode gradient magnetic field, frequency encode gradient magnetic field, echo signal, and sampling period, respectively. FIG. 4 shows a sequence chart for two repetitions, in which the first half is suffixed with “-1” and the second half is suffixed with “-2”. Then, within one repetition time (TR), the 90 ° excitation RF pulse 401 and the slice selection gradient magnetic field pulse 403 are applied almost simultaneously to excite the spin in the desired imaging region, and the encode gradient magnetic field pulse in the phase encoding direction. 405 and an encoded gradient magnetic field pulse 406 in the frequency encoding direction are applied to encode position information into the phase of the excited spin, and a 180 ° refocus RF pulse 402 and a slice selective gradient magnetic field pulse 404 are applied substantially simultaneously. Then, the phase of the spin is refocused to form an echo signal, and the echo signal is measured in the sampling period 410.

In such Spin Echo sequence, test gradient magnetic fields 407 and 408 having the same applied amount are applied before and after the refocus RF pulse 402 in the physical axis direction (frequency encoding direction in the example of FIG. 4) for measuring the error magnetic field. An echo signal superimposed with an error magnetic field (due to eddy current or mechanical vibration) induced by the test gradient magnetic field so that the test gradient magnetic field 408 becomes zero immediately before the time TE, or immediately after or Sampling after

In the error magnetic field measurement sequence as described above, the polarities of the test gradient magnetic fields 407 and 408 are made different in order to exclude the influence of the error magnetic field caused by the gradient magnetic fields 403, 404, 405, 406, and 407 for two-dimensional imaging and the influence of static magnetic field inhomogeneity. (I.e. reversed) or with or without a test gradient magnetic field. For this reason, in another repetition time (TR), the polarities of the test gradient magnetic fields 407 and 408 are reversed, and the same except for the test gradient magnetic fields 407 and 408. In the test gradient magnetic fields 407 and 408, measurement of positive polarity is Scan (+), and measurement of negative polarity is Scan (-). Alternatively, measurement with a test gradient magnetic field may be Scan (+), measurement without a test gradient magnetic field may be Scan (-), and vice versa.

For echo signal measurement, as shown in Fig. 5, Scan (+) and Scan (-) are sequentially performed for each phase encode value, and this is repeated Np * Nf times as necessary for biaxial phase encoding of the excitation cross section. It is. Note that the polarity inversion of the test gradient magnetic field does not have to be alternate. After the measurement of one of the test gradient magnetic fields is completed, the polarity of the test gradient magnetic field is inverted and the measurement may be repeated.

The error magnetic field measurement sequence in FIG. 4 does not need to be excited by an RF pulse after applying the test gradient magnetic field as in the error magnetic field measurement sequence of Patent Document 3, so during the gradient magnetic field application or immediately after the gradient magnetic field is made zero. From this, there is an advantage that the error magnetic field can be measured.

The frequency resolution for the vibration error magnetic field in the above two methods is determined by the echo signal sampling bandwidth (BW), and the measurement capability on the low frequency side is determined by the echo signal acquisition time (window time).

The error magnetic field measurement unit 201 generates the error magnetic field measurement sequence of Patent Document 3 or FIG. 4 described above, and causes the measurement control unit 6 to execute the measurement of the echo signal on which the error magnetic field is superimposed.

(Acquisition of time-series phase image)
Next, details of acquisition of time-series phase image data in step 302 will be described based on the flowchart shown in FIG. The time-series phase image data acquisition process is performed in both of the two methods (low frequency component measurement and high frequency component measurement) described in the above-described vibration error magnetic field measurement.

In step 651, the error magnetic field image acquisition unit 202 uses the echo signal measured in step 301 as a three-dimensional data set S + (kx, for each of Scan (+) and Scan (-) as indicated by 601 and 602 in FIG. ky, ti) and S- (kx, ky, ti). Here, kx and ky are spatial frequencies in the frequency encoding direction and the phase encoding direction, respectively, and ti indicates discrete time points (i = 1, 2,..., N) of the sampled echo signal. The time interval in the ti direction (time resolution) is uniquely determined by the sampling frequency, and the number of data is uniquely determined by the frequency direction resolution Freq. #.

In step 652, the error magnetic field image acquisition unit 202 performs two-dimensional Fourier transform separately at each time ti using the two data sets S +, S- as variables kx, ky, and the two-dimensional complex image I + at each time. Obtain (x, y, ti) and I- (x, y, ti). Using this two-dimensional complex image, phase images φ + (x, y, ti) and φ− (x, y, ti) at each time ti are created (603, 604).

In step 653, the error magnetic field image acquisition unit 202 outputs the phase image data set φ (x, y, excluding the influence of the error magnetic field caused by the gradient magnetic fields 403, 404, 405, 406, and 407 for imaging and the influence of the static magnetic field inhomogeneity. In order to create (ti) (605), the difference between the two phase image data φ + (x, y, ti), φ− (x, y, ti) is taken.

Phase image data when the test gradient magnetic field is reversed and measured twice (bipolar measurement)
φ (x, y, ti) = [φ + (x, y, ti) -φ- (x, y, ti)] / 2
In the case of unipolar measurement,
φ (x, y, ti) = φ + (x, y, ti) -φ- (x, y, ti)
And FIG. 6 shows the case of bipolar measurement. The phase image data thus obtained reflects the phase based on the vibration error magnetic field accompanied by attenuation due to the vibration of the MRI apparatus structure induced by the application of the gradient magnetic field.

Finally, based on the fact that the phase of the phase image and the magnetic field intensity are in a proportional relationship, the vibration error magnetic field image data Be (x, y, ti) can be obtained as follows.

Be (x, y, ti) = φ (x, y, ti) / (γti)
Here, γ is a magnetic rotation ratio. If this calculation is performed for each time ti, vibration error magnetic field image data representing an error magnetic field distribution for each sampling time can be acquired.

Up to the above is the process of acquiring the vibration error magnetic field data for each sampling time, but the above two methods (low frequency component measurement and high frequency component measurement), the error magnetic field is converted into two-dimensional or three-dimensional spatial information. It is possible to obtain. However, when such detailed spatial information is not required, it is only necessary to measure the error magnetic field in specific spatial coordinates as shown in Patent Document 3 or Patent Document 4.

(Analysis of vibration error magnetic field)
Next, details of the vibration error magnetic field analysis in step 303 will be described.

The error magnetic field calculation unit 303 performs an analysis to obtain a frequency distribution for each spatial component of the vibration error magnetic field data acquired by the vibration error magnetic field measurement described above based on the flowchart shown in FIG. Hereinafter, the processing of each step will be described in detail.

In step 701, the error magnetic field calculation unit 303 calculates the vibration error magnetic field Be (x, y, ti) at each time ti acquired in step 653 for each direction (X, Y, Z) in which the test gradient magnetic field is applied. , Each term of the spherical harmonic function (hereinafter referred to as the spherical harmonic term). That is, the vibration error magnetic field Be (x, y, ti) at all times ti is decomposed into spherical harmonic terms as follows.

Be (x, y, ti) =
ζ _0,0 (ti) + ζ _{1, -1} (ti) y + ζ _1,0 (ti) z + ζ _1,1 (ti) x + ζ _{2, -2} (ti) xy + ζ _{2,- 1} (ti) yz
+ ζ _2,0 (ti) (3z ² -1) + ζ _2,1 (ti) xz + ζ _2,2 (ti) (x ² -y ² ) + ・・
Here, ζ _{l, m} (ti) (I, m = 0, ± 1, ± 2,...) Is a coefficient value at time ti of each term. The reason for the decomposition is that the correction magnetic field generator for correcting the vibration error magnetic field and the gradient magnetic field coil (first order term) generate a spatially changing magnetic field corresponding to the spherical harmonic term. Therefore, it is desirable that the order of the spherical harmonic term is matched with the specifications of the correction magnetic field generator and the gradient magnetic field coil (first order term) included in the MRI apparatus to be corrected.

On the other hand, when a vibration error magnetic field in specific spatial coordinates as shown in Patent Document 3 or Patent Document 4 is measured, it is decomposed into terms that can be derived by the respective methods.

An effective and desirable decomposition example is to decompose the vibration error magnetic field into a zero-order term and all the first-order terms and correct the vibration error magnetic field for each spherical harmonic term. For example, when a test gradient magnetic field is applied to the X, Y, and Z axes, this corresponds to calculating the spherical harmonic term shown in the following table.

In step 702, the error magnetic field calculation unit 303 arranges the spherical harmonic term data of the vibration error magnetic field calculated in step 701 in the time axis direction and multiplies the window function w (k) in the time axis direction. That is, ζ _{l, m} (t _k ) ← W (tk) ζ _{l, m} (t _k )
Each spherical harmonic term data thus multiplied by the window function is used for the subsequent analysis.

A definition formula of a desirable window function “Tukey Window” in the present embodiment is shown below.

k: integer L = N + 1

Here, L (Window Length) is the same size (length) as the target analysis data, and α (0 ≦ α ≦ 1) is optimized by the target data. FIG. 8 (a) shows an example of Tukey Window, and FIG. 8 (b) shows an example of multiplying the (X to X) term of the vibration error magnetic field by Tukey Window.

In addition, when there are multiple vibration error magnetic field data in the time direction by shifting the signal acquisition start time etc., the overlapping part along the time axis acquires the addition average etc. and combines it into one data Shall.

Next, the error magnetic field calculation unit 303 performs frequency analysis by the following processes of Steps 703 to 705.
In step 703, the error magnetic field calculation unit 303 performs discrete Fourier transform for each spherical harmonic term data of the vibration error magnetic field multiplied by the window function in step 702, together with an appropriate zero padding process.

In step 704, the error magnetic field calculation unit 303 creates an absolute value spectrum and a phase spectrum from the complex spectrum obtained by the discrete Fourier transform. In order to calculate the phase value from the complex number data, a known technique may be used. An unwrapping process may be applied to the phase spectrum. FIG. 9 (a) shows an example of an absolute value spectrum regarding the Y to X component, and FIG. 9 (b) shows an example of a phase spectrum.

In step 705, the error magnetic field calculation unit 303 performs frequency peak analysis on the frequency spectrum acquired in step 704. Therefore, the fact that the Lorentzian function in the frequency domain (Freq. Domain) and the damped oscillation function in the time domain (Time Domain) are mathematically associated through Fourier transformation is used. Therefore, a damped oscillation function in the time domain corresponding to the parameter value of the Lorentzian function obtained by approximating (that is, fitting) the spectrum shape near the frequency peak in the frequency domain is obtained.

A Lorentzian function with a frequency ≦ f as a variable is defined by the following equation.

This function has the following correspondence relationship with the time domain by using Fourier transform.

Based on this relationship, the error magnetic field calculation unit 303 performs frequency peak analysis by fitting a Lorentzian function to the absolute value spectrum (fitting). When there are a number of significant frequency peaks, nonlinear approximation may be performed using the mathematically known Levenberg-Marquardt method or Nelder-Mead method.

The frequency peak analysis may use a Gaussian function as shown below instead of the Lorentzian function. It is desirable that these are properly used according to the vibration characteristics.

When the identification of the frequency peak is completed, the phase value at the peak frequency is read from the phase spectrum obtained in Step 703-2. Since the phase spectrum is a discrete value, the accuracy may be improved by appropriately using complementary processing together.

The parameter values obtained by the above frequency analysis, such as the damping constant τ, the peak frequency f ₀ , and the phase, are stored in the storage unit such as the magnetic disk 23 as characteristic values representing the error magnetic field, and the vibration in step 704 is performed. When the output value of the correction magnetic field for correcting the error magnetic field is obtained, it is read out and used for the calculation.

(Calculation of correction magnetic field)
Next, details of the correction magnetic field calculation for correcting the vibration error magnetic field in step 304 will be described.

The error magnetic field correction unit 204 uses the parameter values representing the time domain damped oscillation function obtained in step 705 to calculate the output value of the correction magnetic field for correcting the vibration error magnetic field according to the input gradient magnetic field waveform. Ask. Since the magnetic field component that corrects the vibration error magnetic field is superimposed on the gradient magnetic field waveform, this error magnetic field correction unit 204 is added to the eddy current correction function (control board) already installed in the measurement control unit 6 of the MRI apparatus. It may be implemented or a new function may be added. From the viewpoint of operation speed, it is desirable to perform calculation by hardware. However, if the processing speed is in time, the calculation processing unit 7 may include the error magnetic field correction unit 204 and perform calculation by software. Details of the correction magnetic field calculation will be described below.

First, the expression (formulation) of the output waveform of the correction magnetic field component will be described. Since the concept is the same for any correction component, the primary gradient component of the vibration error magnetic field will be described as a representative.
The impulse response function VGC ^j (dl ^j (s), t) for the input gradient magnetic field waveform (differential value of the gradient magnetic field waveform on the j-axis) dl ^j (S) is modeled as follows.

When t = 0

It becomes. here,

It is.

Based on this model, in order to obtain the output of the correction component for an input waveform of an arbitrary gradient magnetic field, the successive response correction for the input waveform (differential value of the gradient magnetic field waveform) is performed in the same manner as the error magnetic field correction caused by the eddy current. Specific processing is shown in the following equation. When the input waveform dl ^j (t _k ) is given on the discontinuous time axis t _k (k = 0, 1, 2,... N), the oscillating magnetic field at time t is given by the following equation.

:
:

(*) Note that this is a vector sum.

As described above, the vibration error magnetic field is calculated at any time, and its real part is output as a correction value.

Re [...]: Real part When the exponential function cannot be processed directly on the hardware side, the following relational expression is used.

(Calculation of correction magnetic field)
Next, details of phase correction of the correction magnetic field in step 305 will be described.
The phase calibration unit 206 obtains a phase calibration value for performing phase calibration of the output of the correction magnetic field obtained in step 704. When the phase calibration unit 206 is mounted in the measurement control unit 6, the phase calibration value obtained by the phase calibration unit 206 is used to perform phase calibration of the correction magnetic field, and the correction magnetic field after the phase calibration is corrected to the correction magnetic field generator and Output to the gradient magnetic field power supply. When the phase calibration unit 206 is mounted in the arithmetic processing unit 7, the measured phase calibration value is notified to the measurement control unit 6. The measurement control unit 6 performs phase calibration on the correction magnetic field calculated in Step 704 using this phase calibration value, and outputs the correction magnetic field after the phase calibration to the correction magnetic field generator and the gradient magnetic field power source.

The purpose of the phase calibration is to cancel the phase error since an error occurs between the output phase of the correction magnetic field obtained in step 704 and the phase at the frequency peak obtained in steps 703 to 705. Therefore, the phase value for each frequency is measured. The relationship between the phase calibration values of these frequencies is stored in advance on a magnetic disk or the like, and the phase calibration unit 206 reads out the phase calibration value according to the frequency peak value obtained in Step 703-3 to obtain the phase of the correction magnetic field. Perform calibration. FIG. 10 shows a graph of an example of phase calibration data. In this graph, the horizontal axis is the set frequency, and the vertical axis is the measurement phase. A Y-intercept with a frequency of zero corresponds to a phase error (offset).

As described above, the MRI apparatus and the vibration error magnetic field reduction method of this embodiment identify the frequency peak in the frequency spectrum of the vibration error magnetic field and the waveform shape in the vicinity of the peak using a predetermined function, and are obtained by the identification. A correction magnetic field having a decay time constant corresponding to the parameter value is obtained according to the input gradient magnetic field waveform, and the obtained correction magnetic field is superimposed on the gradient magnetic field and output. This makes it possible to correct not only the eddy current magnetic field but also the vibration error magnetic field accompanying the vibration of the mechanical structure of the MRI apparatus accompanying the application of the gradient magnetic field, and as a result, the image quality can be improved.

Finally, FIG. 11 shows an example of the effect when the vibration error magnetic field reduction of the present invention is actually applied. FIG. 11 is an image obtained by performing multi-slice imaging of an axial cross section of a phantom using an FSE sequence (TR = 733 msec, IET = 8 msec, phase encoding direction = vertical direction). It is an image obtained when the vibration error magnetic field reduction processing is not performed, and (b) is an image obtained when the vibration error magnetic field reduction processing of the present invention is performed. By applying the vibration error magnetic field reduction processing of this embodiment, it is understood that image shading that occurs at both ends in the left-right direction is eliminated.

As understood from the above description of the embodiments of the present invention, the present invention has the following features. That is, the MRI apparatus of the present invention is
A static magnetic field generator for generating a static magnetic field in the imaging space, a gradient magnetic field generator for generating a gradient magnetic field superimposed on the static magnetic field, and a correction for correcting an error magnetic field generated in the imaging space due to the application of the gradient magnetic field A correction magnetic field generation unit for generating a magnetic field, a static magnetic field generation unit, a gradient magnetic field generation unit, a structure for supporting the correction magnetic field generation unit and supporting them, and a target disposed in the imaging space based on a predetermined pulse sequence. A measurement control unit that measures an echo signal from a specimen, and a correction magnetic field calculation unit that obtains a correction magnetic field for correcting an error magnetic field generated in an imaging space with application of a gradient magnetic field,
The correction magnetic field calculation unit obtains an error magnetic field including a vibration error magnetic field based on the vibration of the structure part caused by application of the gradient magnetic field, and obtains a correction magnetic field for correcting the obtained error magnetic field.

Preferably, a storage unit is provided for storing a characteristic value representing the error magnetic field obtained by the correction magnetic field calculation unit, and the correction magnetic field calculation unit obtains the correction magnetic field based on the stored characteristic value of the error magnetic field.

Preferably, the correction magnetic field calculation unit includes an error magnetic field measurement unit that causes the measurement control unit to measure an echo signal using a pulse sequence having a test gradient magnetic field, and an error for each sampling time using the echo signal. An error magnetic field image acquisition unit that acquires error magnetic field image data representing a magnetic field distribution, an error magnetic field calculation unit that calculates a parameter value of a damped oscillation function that represents an error magnetic field using the error magnetic field image data, and a calculated parameter value And a correction magnetic field calculation unit for calculating a correction magnetic field.

Preferably, the error magnetic field measurement unit causes the measurement control unit to measure an echo signal on which information on the eddy current error magnetic field and the vibration error magnetic field is superimposed, thereby obtaining the information on the eddy current error magnetic field and the vibration error magnetic field. Acquire together.

Preferably, the pulse sequence applies an encoded gradient magnetic field pulse in at least two axial directions, and the error magnetic field image acquisition unit obtains an echo signal by Fourier transform in at least two axial directions at every sampling time. Error magnetic field image data for each sampling time is obtained using the phase image data.

Preferably, the error magnetic field measuring unit causes the measurement control unit to measure the echo signal by changing the test gradient magnetic field, and the error magnetic field image acquiring unit is configured to output the phase image data having a different test gradient magnetic field for each sampling time. Error magnetic field image data for each sampling time is obtained using the difference.

Preferably, the error magnetic field calculation unit decomposes each error magnetic field image data into a plurality of spherical harmonic terms, calculates a parameter value of a damped oscillation function representing the error magnetic field for each spherical harmonic term,
The correction magnetic field calculation unit calculates the correction magnetic field for each spherical harmonic term based on the parameter value of the damped oscillation function representing the error magnetic field for each spherical harmonic term.

Preferably, the error magnetic field calculation unit applies a Lorentzian function or a Gaussian function to a spectral distribution obtained by performing Fourier transform of the error magnetic field image data in the time axis direction for each spherical harmonic term, and thereby attenuated vibration representing the error magnetic field. Calculate function parameter values.

Preferably, the error magnetic field calculation unit applies a Lorentzian function or a Gaussian function to a waveform near a frequency peak in a spectrum distribution in the frequency domain, and a time domain damped oscillation function corresponding to a Fourier transform of the Lorentzian function or Gaussian function. The parameter value including the decay time constant is calculated.

Preferably, the correction magnetic field calculation unit creates a model of an impulse response function for the gradient magnetic field waveform using the parameter value of the damped oscillation function, and corrects the magnetic field as a sequential response of the model for any input gradient magnetic field waveform. Is calculated.

Preferably, a phase calibration unit for calibrating the phase error between the phase of the correction magnetic field and the phase at the frequency peak of the spectrum distribution is provided.

Also preferably, the pulse sequence includes a test gradient magnetic field before and after the refocus RF pulse, and measures an echo signal after the test gradient magnetic field after the refocus RF pulse becomes zero.

The vibration error magnetic field reduction method of the present invention is a vibration error magnetic field reduction method for correcting a vibration error magnetic field based on vibration of a structure of a magnetic resonance imaging apparatus caused by application of a gradient magnetic field using a correction magnetic field. A measurement step for measuring an echo signal using a pulse sequence having a gradient magnetic field, a step for acquiring error magnetic field image data representing an error magnetic field distribution for each sampling time using the echo signal, and an error magnetic field for each sampling time A parameter value calculating step for calculating a parameter value of a damped oscillation function representing a vibration error magnetic field using image data, and a correction magnetic field for calculating a correction magnetic field based on the parameter value of the damped oscillation function representing the calculated vibration error magnetic field And a calculating step.

Preferably, the parameter value calculating step uses a Lorentzian function or a Gaussian function for a spectral distribution obtained by Fourier transforming the error magnetic field image data for each sampling time in the time axis direction, and a damped oscillation function representing the error magnetic field. The parameter value of is calculated.

Preferably, the parameter value calculating step obtains a time domain damped oscillation function corresponding to the Lorentzian function or Gaussian function applied to the spectrum distribution in the frequency domain, and sets the parameter value including the damping time constant of the damped oscillation function. calculate.

Preferably, the correction magnetic field calculation step creates a model of an impulse response function for the gradient magnetic field waveform using the parameter value of the damped oscillation function, and uses the correction magnetic field as a sequential response of the model for any input gradient magnetic field waveform. Is calculated.

1 Gradient magnetic field generation system, 2 Static magnetic field generation system, 3 Transmission system, 4 Operation unit, 5 Reception system, 6 Measurement control unit, 7 Signal processing system, 8 Arithmetic processing unit, 9 Subject, 10 Gradient magnetic field coil, 11 Gradient Magnetic field power supply, 12 RF oscillator, 13 modulator, 14 RF amplifier, 15 RF transmitter coil, 16 RF receiver coil, 17 signal amplifier, 18 quadrature phase detector, 19 A / D converter, 20 ROM, 21 RAM, 22 light Magnetic disk, 23 display, 24 magnetic disk, 25 trackball or mouse, 26 keyboard

Claims

A static magnetic field generator for generating a static magnetic field in the imaging space;
A gradient magnetic field generator for generating a gradient magnetic field superimposed on the static magnetic field;
A correction magnetic field generator for generating a correction magnetic field for correcting an error magnetic field generated in the imaging space due to the application of the gradient magnetic field;
The static magnetic field generation unit, the gradient magnetic field generation unit, and the correction magnetic field generation unit are installed and support the structure,
A measurement control unit that measures an echo signal from a subject arranged in the imaging space based on a predetermined pulse sequence;
A correction magnetic field calculation unit for obtaining a correction magnetic field for correcting an error magnetic field generated in the imaging space with the application of the gradient magnetic field;
A magnetic resonance imaging apparatus comprising:
The correction magnetic field calculation unit obtains an error magnetic field including a vibration error magnetic field based on the vibration of the structure part caused by application of the gradient magnetic field, and obtains the correction magnetic field for correcting the obtained error magnetic field. Magnetic resonance imaging device.
In the magnetic resonance imaging apparatus according to claim 1,
A storage unit for storing a characteristic value representing an error magnetic field obtained by the correction magnetic field calculation unit;
The magnetic resonance imaging apparatus, wherein the correction magnetic field calculation unit obtains the correction magnetic field based on the stored characteristic value of the error magnetic field.
In the magnetic resonance imaging apparatus according to claim 1,
The correction magnetic field calculation unit
An error magnetic field measurement unit that causes the measurement control unit to measure an echo signal using a pulse sequence having a test gradient magnetic field;
Using the echo signal, an error magnetic field image acquisition unit that acquires error magnetic field image data representing an error magnetic field distribution for each sampling time, and
An error magnetic field calculation unit for calculating a parameter value of a damped oscillation function representing the error magnetic field using the error magnetic field image data;
A correction magnetic field calculation unit for calculating the correction magnetic field based on the calculated parameter value;
A magnetic resonance imaging apparatus comprising:
In the magnetic resonance imaging apparatus according to claim 3,
The error magnetic field measurement unit causes the measurement control unit to measure an echo signal in which the information on the eddy current error magnetic field and the vibration error magnetic field is superimposed, thereby obtaining both the information on the eddy current error magnetic field and the vibration error magnetic field. A magnetic resonance imaging apparatus characterized by comprising:
In the magnetic resonance imaging apparatus according to claim 3,
The pulse sequence applies an encoded gradient magnetic field pulse in at least two axial directions,
The error magnetic field image acquisition unit obtains error magnetic field image data for each sampling time using phase image data obtained by performing Fourier transform on the echo signal in the at least two axial directions for each sampling time. Magnetic resonance imaging apparatus.
In the magnetic resonance imaging apparatus according to claim 3,
The error magnetic field measurement unit makes the measurement control unit measure the echo signal by changing the test gradient magnetic field,
The magnetic resonance imaging apparatus, wherein the error magnetic field image acquisition unit obtains error magnetic field image data for each sampling time using a difference between phase image data having different test gradient magnetic fields for each sampling time.
In the magnetic resonance imaging apparatus according to claim 3,
The error magnetic field calculation unit decomposes each error magnetic field image data into a plurality of spherical harmonic terms, calculates a parameter value of a damped oscillation function representing the error magnetic field for each spherical harmonic term,
The magnetic resonance imaging apparatus, wherein the correction magnetic field calculation unit calculates the correction magnetic field for each spherical harmonic term based on a parameter value of a damped oscillation function representing an error magnetic field for each spherical harmonic term.
The magnetic resonance imaging apparatus according to claim 7,
The error magnetic field calculation unit applies a Lorentzian function or a Gaussian function to a spectral distribution obtained by Fourier transforming the error magnetic field image data in the time axis direction for each spherical harmonic term, and a damped oscillation function representing the error magnetic field. The parameter value of the magnetic resonance imaging apparatus is calculated.
The magnetic resonance imaging apparatus according to claim 8,
The error magnetic field calculation unit applies a Lorentzian function or a Gaussian function to a waveform near a frequency peak in the spectrum distribution in the frequency domain, and attenuates a time domain damped oscillation function corresponding to a Fourier transform of the Lorentzian function or Gaussian function. A magnetic resonance imaging apparatus characterized by calculating a parameter value including a constant.
In the magnetic resonance imaging apparatus according to claim 3,
The correction magnetic field calculation unit creates a model of an impulse response function with respect to a gradient magnetic field waveform using the parameter value of the damped oscillation function, and uses the correction magnetic field as a sequential response of the model with respect to an arbitrary input gradient magnetic field waveform. A magnetic resonance imaging apparatus characterized by calculating.
The magnetic resonance imaging apparatus according to claim 9,
A magnetic resonance imaging apparatus comprising: a phase calibration unit for calibrating a phase error between a phase of the correction magnetic field and a phase at a frequency peak of the spectrum distribution.
In the magnetic resonance imaging apparatus according to claim 3,
The pulse sequence includes the test gradient magnetic field before and after the refocus RF pulse, and measures the echo signal after the test gradient magnetic field after the refocus RF pulse becomes zero. Magnetic resonance imaging device.
A vibration error magnetic field reduction method for correcting a vibration error magnetic field based on vibration of a structure of a magnetic resonance imaging apparatus caused by application of a gradient magnetic field using a correction magnetic field,
A measurement step of measuring an echo signal using a pulse sequence having a test gradient magnetic field;
Using the echo signal, obtaining error magnetic field image data representing an error magnetic field distribution for each sampling time; and
A parameter value calculating step for calculating a parameter value of a damped oscillation function representing the vibration error magnetic field using the error magnetic field image data for each sampling time;
A correction magnetic field calculation step for calculating the correction magnetic field based on a parameter value of a damped vibration function representing the calculated vibration error magnetic field;
A vibration error magnetic field reduction method characterized by comprising:
In the vibration error magnetic field reduction method according to claim 13,
The parameter value calculating step uses a Lorentzian function or a Gaussian function for a spectral distribution obtained by Fourier transforming the error magnetic field image data for each sampling time in the time axis direction, and using the Lorentzian function or the Gaussian function, a damped oscillation function representing the error magnetic field. A parameter value of the vibration error magnetic field reduction method is calculated.
In the error magnetic field correction method according to claim 14,
The parameter value calculating step obtains a time domain damped oscillation function corresponding to the Lorentzian function or Gaussian function applied to the spectral distribution in the frequency domain, and calculates a parameter value including a damping time constant of the damped oscillation function. A method for reducing a vibration error magnetic field.
In the vibration error magnetic field reduction method according to claim 14,
The correction magnetic field calculation step creates a model of an impulse response function for a gradient magnetic field waveform using the parameter value of the damped oscillation function, and uses the correction magnetic field as a sequential response of the model for an arbitrary input gradient magnetic field waveform. A vibration error magnetic field reduction method characterized by calculating.